AI & Automation Statement
DJF IT Cyberworks Ltd designs AI-assisted systems and business automation to improve workflows, reduce repetitive work and help organisations use technology more effectively. We believe automation should support people, not remove appropriate human judgement, accountability or control.
01 // Our Approach to AI
We use artificial intelligence and automation as practical business tools rather than as replacements for responsible human decision-making.
The objective of an AI or automation project should be clear: solve a genuine business problem, improve a workflow, reduce repetitive administration or provide useful assistance to users.
People remain responsible for important business decisions and the appropriate use of automated output.
Automation should be designed around a defined workflow or business requirement, not added simply because AI is available.
Systems should process only the information reasonably necessary for their intended function.
Higher-impact actions should include appropriate safeguards, limits or human review.
02 // What We Mean by AI & Automation
Depending on the project, AI and automation may include:
- AI-powered website assistants and virtual advisors.
- Automated quotation and estimation workflows.
- Intelligent form processing.
- Automated email or notification workflows.
- Business process automation.
- API-connected systems.
- Document and information processing.
- AI-assisted classification, summarisation or content generation.
- Internal workflow assistants.
- Rule-based and AI-assisted decision-support tools.
03 // Human-Driven by Design
DJF IT's approach is based on human-driven automation.
AI can analyse information, suggest actions, generate content or automate routine steps, but important decisions should remain subject to appropriate human oversight.
Where a workflow could have a significant effect on a person, business, financial position, legal position or safety-related outcome, AI output should not be treated as automatically correct simply because it was produced by an automated system.
04 // AI Output Is Not Infallible
AI systems can produce useful and highly capable output, but they can also produce information that is incomplete, inaccurate, outdated or inappropriate for a particular context.
Automated output should therefore be reviewed in proportion to the importance of the decision or task.
Greater levels of human verification should be applied where the consequences of an error would be more significant.
05 // No Professional Advice
Unless expressly agreed as part of a specific service, AI-generated content supplied through DJF IT systems is intended for information, assistance or workflow-support purposes.
It should not automatically be treated as a substitute for qualified legal, medical, financial, accounting, engineering, safety or other regulated professional advice.
06 // Data Minimisation
AI systems should be designed so that they receive only information reasonably required to perform their intended function.
Users should avoid entering passwords, payment-card details, authentication secrets or unnecessary sensitive personal information into public AI interfaces.
Where possible, data collection and prompt inputs should be limited to what is relevant to the intended business workflow.
07 // Confidential Information
Confidential, proprietary or commercially sensitive information should be handled according to the requirements of the relevant project.
Clients should not submit information to an AI system unless they are authorised to use that information for the relevant purpose.
Where a project involves particularly sensitive information, additional technical, contractual or processing arrangements may be appropriate.
08 // API Keys, Credentials & Privileged Access
Sensitive API credentials, private authentication tokens and administrative secrets should not be exposed in publicly accessible frontend code where avoidable.
Privileged operations should be handled through appropriate trusted or server-side processes where the architecture of the project requires it.
AI interfaces should not be given unrestricted access to systems or information unless such access is genuinely required and appropriately controlled.
09 // Access & Abuse Controls
Public AI and automation systems may require technical limits designed to reduce misuse, excessive automated requests, resource abuse or attempts to bypass normal system behaviour.
Depending on the project, controls may include:
- Request or usage limits.
- Input-length restrictions.
- Session controls.
- Server-side validation.
- Restricted tool or API access.
- Abuse monitoring.
- Blocking of unsupported or unsafe requests.
10 // Third-Party AI Providers
Some AI solutions may use models, APIs, cloud infrastructure or specialist services supplied by third-party technology providers.
These providers operate their own systems and may have their own service terms, privacy practices, availability arrangements and technical limitations.
The appropriate provider and architecture will depend on the requirements, sensitivity and intended use of the project.
11 // Automation Boundaries
Automation should have a clearly defined scope.
A system designed to prepare a quotation, classify an enquiry or send a routine notification should not automatically be treated as authorised to perform unrelated actions.
Where practical, automated systems should be limited to the permissions, data and actions required for their intended workflow.
12 // High-Impact Decisions
DJF IT does not promote the use of general-purpose AI as an unquestioned final authority for high-impact decisions.
Appropriate human involvement should be maintained where an automated output could materially affect:
- Employment or recruitment.
- Significant financial decisions.
- Legal rights or obligations.
- Health or medical matters.
- Safety-critical activities.
- Access to important services.
- Other decisions with significant consequences.
13 // Testing & Deployment
AI and automation systems should be tested in a manner proportionate to their intended use before production deployment.
Testing may include reviewing expected inputs, unusual inputs, error behaviour, integrations, permissions and failure scenarios.
Because AI behaviour can vary depending on context, models, providers and input, testing reduces risk but cannot guarantee identical output in every situation.
14 // Monitoring & Improvement
AI-assisted systems may require review after deployment as user behaviour, business requirements, providers or underlying technologies change.
Where an active support arrangement applies, appropriate improvements may include:
- Updating prompts or business rules.
- Adjusting workflow logic.
- Improving safeguards.
- Updating integrations.
- Responding to provider or API changes.
15 // Client Responsibilities
Clients remain responsible for determining whether an AI or automation solution is suitable for their intended business use.
Clients should ensure that:
- Information supplied to the system may lawfully be used for the intended purpose.
- Relevant employees or users understand the role of the automated system.
- Important output is reviewed where appropriate.
- Credentials and administrative access are protected.
- Changes in business requirements are communicated where they affect the operation of the system.
16 // Intellectual Property & AI Content
AI-assisted content may involve outputs generated with the assistance of third-party AI technologies.
Clients should not assume that every AI-generated output is automatically exclusive, original or free from potential third-party considerations.
Where intellectual property is commercially important, appropriate human review and project-specific arrangements may be required.
17 // Privacy & Data Protection
AI and automation projects may involve the processing of personal information.
Our wider approach to personal data is explained in our Privacy Policy .
Our broader technical security approach is explained in our Data Protection & Security statement.
18 // AI Does Not Mean Full Autonomy
The presence of AI within a system does not mean that the system should operate without limits, supervision or accountability.
In many business environments, the most effective solution combines:
- Automated processing.
- Business rules.
- API integrations.
- AI assistance.
- Human judgement.
DJF IT designs technology around the workflow rather than forcing a workflow around the technology.
IMPORTANT: AI and automation can improve efficiency and provide valuable assistance, but no AI model or automated system should be assumed to be completely accurate, available or suitable for every situation. Appropriate human oversight remains essential where the consequences of error are significant.
19 // Review & Updates
AI technology develops quickly. This statement may therefore be reviewed and updated as our services, technology, business practices or relevant requirements change.
The latest version will be published on this page together with its revision date.
20 // AI & Automation Enquiries
If you would like to discuss an AI assistant, automation workflow, API-connected business process or custom digital system, contact:
Company Information
DJF IT CYBERWORKS LTD
16760496
England and Wales
34 Selworthy Drive
Crewe
England
CW1 3RR
